Hi,
what is your oppinion on the way messages are passing through. Should
they be unmodified or mangled them in order to try to remove
potentially problematic characters?
At the moment syslogd removes potentially problematic characters due
to the security implication:
http://lkml.org/lkml/2006/8/26/87
http://lkml.org/lkml/2006/8/25/408
However, this breaks messages encoded in UTF-8 and with more machines
running in a non-POSIX locale more log messages will be emitted in
different locales and character encodings as well. Thus, this may not
be practical.
However, this is the current behaviour in the unreleased CVS version.
The relevant patch is 1.29 for syslogd.c.
Regards,
Joey
-- We all know Linux is great... it does infinite loops in 5 seconds. -- Linus TorvaldsReceived on Wed May 23 2007 - 18:25:55 CEST
This archive was generated by hypermail 2.2.0 : Wed May 23 2007 - 18:27:21 CEST